如何在Lex中配置槽位接收任意字符串？反馈场景槽位匹配咨询

Hey there! Let’s tackle your Amazon Lex questions with practical, hands-on solutions based on real-world usage:

1. How to set up a slot to accept arbitrary strings in Lex? #

If you’re working with Amazon Lex V2, the simplest solution is to use the built-in AMAZON.FreeText slot type. This type is purpose-built for capturing unstructured, open-ended text—no restrictions, no predefined values needed. Just assign this slot type to your target slot, and it will capture whatever the user types or says.

For those still maintaining Lex V1 bots, you’ll need a custom slot setup:

• Create a new custom slot type, but leave the enumeration values list empty.
• When configuring the slot in your intent, set the Value Constraints option to "None". This tells Lex not to limit inputs to specific values, allowing it to accept any string the user provides.

2. Slot type confusion for feedback collection: Can I use Regex, or are there workarounds? #

Feedback collection is a super common use case, and Lex has you covered with multiple options:

Using Regex for Slot Matching #

Absolutely! Lex V2 supports regex-based custom slot types, which is perfect if you want to enforce basic structure on feedback (like minimum length, excluding special characters, etc.). Here’s how to set it up:

1. Create a new custom slot type.
2. Under "Slot type source", select Regular expression.
3. Input your desired regex pattern (for example, .{5,} to ensure feedback is at least 5 characters long—tweak this to fit your requirements).
4. Assign this custom slot type to your feedback slot, and Lex will only capture inputs that match your regex rules.

Alternative Workarounds #

If regex feels too rigid, or you want maximum flexibility for open-ended feedback:

• Stick with AMAZON.FreeText: This is the go-to for unrestricted feedback, as it lets users share whatever they want without limitations.
• Split into structured slots: If you need both quantitative and qualitative feedback (like a star rating + written comment), split the intent into two slots: use AMAZON.Number for the rating, and AMAZON.FreeText for the detailed comment. This keeps data organized while still letting users provide rich feedback.
• Guide users with prompts: Add dialogue prompts to steer users (e.g., "Could you tell us a bit more about your experience?")—this helps encourage meaningful feedback without restricting input.

内容的提问来源于stack exchange，提问作者Ankita Prasad